> in a sense that he is talking about a seeded pnrg

That little "p" makes all the difference. The post I was replying to was asking about the precise definition of "random", not "pseudorandom".

As you allude to, any real randomness in such a scheme needs to come in the form of a seed.